A method and system for early warning and control of decarburization in a steel rolling process
By constructing a decarbonization early warning model, and performing decarbonization prediction and graded early warning based on real-time parameters, the problems of lag and error in decarbonization control in steel rolling production were solved, and precise and intelligent decarbonization process control was achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-05-19
- Publication Date
- 2026-07-31
AI Technical Summary
The lack of systematic decarburization prediction and early warning methods in existing steel rolling production leads to delayed response and large judgment errors, making it difficult to achieve precise control of the decarburization process.
A decarbonization early warning model is constructed using a multivariate nonlinear regression algorithm. Based on real-time collected parameters such as heating temperature, heating time, and air-fuel ratio, decarbonization prediction is performed. Process adjustment instructions are generated through graded early warning judgment to achieve closed-loop control.
It enables accurate prediction and timely early warning of decarburization in the steel rolling process, reduces judgment errors, improves the level of intelligence and standardization in production, and ensures effective control of the decarburization process.
Smart Images
Figure CN122480097A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of steel rolling production process control technology, and in particular to a method and system for early warning and control of decarburization in the steel rolling process. Background Technology
[0002] In the steel rolling process, the degree of decarburization of the rolled material directly affects the mechanical properties and surface quality of the product, and is one of the core quality control indicators in steel rolling production. The decarburization of rolled material is mainly affected by various parameters such as the heating time, heating temperature, and heating atmosphere (air-fuel ratio) of the heating furnace. If these process parameters are not properly controlled, it is very easy to cause the decarburization of the rolled material to exceed the standard, requiring additional remedial processes such as peeling, and increasing production costs.
[0003] In current steel rolling production, decarburization control largely relies on operators' experience for manual judgment and adjustment, lacking systematic decarburization prediction and early warning methods. This results in problems such as response lag and large judgment errors. Furthermore, the lack of a precise correlation between process parameters and decarburization results makes it impossible to predict and dynamically control the decarburization process in advance, thus hindering the effective avoidance of excessive decarburization risks. Therefore, there is an urgent need for a control system capable of accurately predicting steel rolling decarburization, providing timely early warnings, and offering targeted solutions to improve the intelligence and precision of decarburization control in steel rolling production. Summary of the Invention
[0004] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide a method and system for early warning and control of decarburization in the steel rolling process.
[0005] To solve the above technical problems, the technical solution of the present invention is as follows: A method for early warning and control of decarburization in steel rolling process, comprising: Obtain the preset heating process parameter thresholds and preset decarburization specification thresholds for the target steel grade; The actual operating parameters of the target steel in the heating furnace are collected in real time and input into a pre-trained decarburization early warning model to generate a real-time decarburization prediction value of the target steel. A graded early warning judgment is performed based on the real-time decarbonization prediction value and the decarbonization standard threshold to generate a decarbonization early warning signal. The graded early warning judgment includes: if the real-time decarbonization prediction value reaches a first preset proportion of the preset decarbonization standard threshold, a risk warning signal is generated; if the real-time decarbonization prediction value is greater than the preset decarbonization standard threshold, an over-limit warning signal is generated. Based on the risk warning signal or the over-limit warning signal and the actual operating parameters, a corresponding process adjustment instruction is generated and output to the corresponding control system to perform closed-loop control of the decarburization process of the target steel.
[0006] As a preferred embodiment of the decarburization early warning and control method for the steel rolling process described in this invention, the training process of the decarburization prediction model includes: Obtain historical rolling production data and corresponding offline decarburization metallographic detection data; Based on the historical rolling production data, historical heating temperature, historical heating duration, historical air-fuel ratio, and historical residual oxygen content are extracted as training feature inputs. The decarburization layer depth in the offline decarburization metallographic detection data is used as the training label output; The decarbonization prediction model is generated by training the input features and the output labels using a multivariate nonlinear regression algorithm.
[0007] As a preferred embodiment of the decarburization early warning and control method for the steel rolling process described in this invention, the actual operating parameters include at least heating temperature, cumulative heating time, and air-fuel ratio.
[0008] As a preferred embodiment of the decarburization early warning and control method for the steel rolling process described in this invention, the step of generating a corresponding process adjustment instruction and outputting it to the corresponding control system based on the risk warning signal or the exceedance warning signal and the actual operating parameters includes: When the above-mentioned over-limit warning signal is generated, and it is detected that the cumulative heating time in the actual operating parameters reaches the upper limit of the preset time threshold and the target steel has not been taken out of the furnace, a cooling and heat preservation process adjustment command is generated. The cooling and heat preservation process adjustment command is output to the heating furnace control system to prompt the operator or automatically execute the cooling and heat preservation operations.
[0009] As a preferred embodiment of the decarburization early warning and control method for the steel rolling process described in this invention, after outputting the cooling and heat preservation process adjustment command to the heating furnace control system, the method further includes: The updated actual operating parameters after the cooling and heat preservation operations are re-acquired and input into the decarburization prediction model to generate an updated decarburization prediction value. If the updated decarburization prediction value is still greater than the preset decarburization standard threshold, the rolling size increment is calculated based on the difference between the updated decarburization prediction value and the preset decarburization standard threshold. A rolling size adjustment command is generated based on the rolling size increment and output to the rolling line control system to adjust the finished product rolling diameter.
[0010] As a preferred embodiment of the decarburization early warning and control method for the steel rolling process described in this invention, after generating the corresponding process adjustment command and outputting it to the corresponding control system, the method further includes: Obtain actual decarburization test data of the target steel after rolling; The actual operating parameters and the actual decarbonization detection data are added to the training dataset, and the parameters of the decarbonization prediction model are periodically iterated and optimized based on the updated training dataset.
[0011] The present invention also provides a decarburization early warning and control system for the steel rolling process, for implementing the above-mentioned method, comprising: The parameter and data acquisition module is used to acquire the heating process parameters of the target steel and the preset decarburization standard threshold, and to collect the actual operating parameters of the target steel in the heating furnace in real time. The actual operating parameters include at least the heating temperature, cumulative heating time and air-fuel ratio. The decarburization prediction module, connected to the parameter and data acquisition module, is used to input the actual operating parameters into the pre-trained decarburization prediction model to generate the real-time decarburization prediction value of the target steel. The early warning judgment module, connected to the decarbonization prediction module, is used to perform graded early warning judgment on the real-time decarbonization prediction value and the preset decarbonization standard threshold to generate a decarbonization early warning signal; the graded early warning judgment includes: if the real-time decarbonization prediction value reaches a first preset proportion of the preset decarbonization standard threshold, a risk warning signal is generated; if the real-time decarbonization prediction value is greater than the preset decarbonization standard threshold, an over-limit early warning signal is generated. The graded handling control module is connected to the early warning judgment module and the parameter and data acquisition module, respectively. It is used to generate corresponding process adjustment instructions and output them to the corresponding control system based on the risk warning signal or the exceedance warning signal and the actual operating parameters.
[0012] The beneficial effects of this invention are: (1) The present invention incorporates a decarbonization early warning model. This model is based on historical production data and decarbonization detection data, and uses a multivariate nonlinear regression algorithm to construct a quantitative correlation formula between heating process parameters and decarbonization layer depth. Compared with the traditional manual experience judgment mode, the present invention can accurately calculate the decarbonization prediction value by collecting parameters such as heating temperature, heating time, air-fuel ratio, and residual oxygen in real time and inputting them into the model, thus transforming qualitative judgment into quantitative prediction and significantly reducing the decarbonization judgment error.
[0013] (2) The system of this invention has the functions of sound and light alarm for decarbonization exceeding the standard and push notification of handling methods. When the predicted decarbonization value reaches the preset proportion of the decarbonization standard threshold (such as 85%), the system pushes the process fine-tuning suggestion; when the predicted value exceeds the standard threshold, the sound and light alarm in the control room is triggered immediately, and the over-standard warning information is displayed in the terminal pop-up window, and the standardized handling plan is pushed at the same time. This mechanism effectively solves the problems of delayed detection by operators and inconsistent handling measures in traditional decarbonization control, and ensures that operators can grasp the situation and take standardized measures as soon as possible when an anomaly occurs.
[0014] (3) This invention establishes multi-dimensional process setting rules and parameter linkage handling rules. The system not only predicts decarburization, but also monitors in real time whether the process parameters of the heating furnace exceed the preset threshold. When the heating time exceeds the threshold, the system immediately prompts the operator to implement cooling and heat preservation measures for the heating furnace; if the model still predicts excessive decarburization after adjustment, the system further prompts the operator to roll the material to a larger size to reserve the decarburization layer removal allowance for subsequent shot blasting and grinding processes. This closed-loop control logic realizes the complete chain of "parameter monitoring - anomaly identification - process adjustment - effect verification - backup handling", which significantly improves the intelligence and standardization level of decarburization process control.
[0015] (4) The present invention has a big data analysis module that collects process parameters and finished product decarbonization detection data daily and adds them to the model training dataset. The decarbonization early warning model is iteratively optimized periodically (e.g., monthly). By comparing the actual detection results with the model predictions, the fitting parameters in the quantitative correlation formula are continuously corrected, so that the prediction accuracy of the model continues to improve with the accumulation of production data. Attached Figure Description
[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0017] Figure 1 The flowchart is a description of the decarburization early warning and control method for the steel rolling process provided by the present invention. Detailed Implementation
[0018] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings.
[0019] Example 1: See Figure 1 This embodiment provides a method for early warning and control of decarburization in the steel rolling process, which specifically includes the following steps: Step S101: Obtain the preset heating process parameter threshold and the preset decarburization specification threshold for the target steel grade.
[0020] Specifically, for GCr15 bearing steel in the furnace, the pre-set threshold for the decarburized layer depth is ≤0.05mm based on the size specifications. Simultaneously, the pre-set heating process control thresholds are: maximum heating section temperature 1180℃, soaking section temperature range 1120±15℃, total heating time ≤180min, air-fuel ratio control range 1:1.02-1:1.08, and residual oxygen content in the furnace target ≤3%. These parameters can be fine-tuned according to the incoming material specifications.
[0021] Step S102: Collect the actual operating parameters of the target steel in the heating furnace in real time, and input them into the pre-trained decarburization early warning model to generate the real-time decarburization prediction value of the target steel.
[0022] Specifically, firstly, real-time data on the heating temperature, cumulative heating time, high-temperature zone duration, air-fuel ratio, and residual oxygen content of the steel billet in the furnace are collected. Then, the collected real-time operating parameters are input into a pre-trained decarburization prediction model specific to GCr15 bearing steel, which calculates and outputs the predicted decarburization value of the steel billet in real time.
[0023] The training process for the decarbonization prediction model is as follows: First, historical rolling production data and corresponding offline decarburization metallographic detection data are acquired. Then, historical heating temperature, historical heating duration, historical air-fuel ratio, and historical residual oxygen content are extracted from the historical rolling production data as training feature inputs, and the decarburization layer depth from the offline decarburization metallographic detection data is used as the training label output. Finally, a multivariate nonlinear regression algorithm is used to train the training feature inputs and training label outputs to fit and generate the decarburization prediction model.
[0024] Step S103: Based on the real-time decarbonization prediction value and the decarbonization standard threshold, a graded early warning judgment is made to generate a decarbonization early warning signal. The graded early warning judgment includes: if the real-time decarbonization prediction value reaches the first preset proportion of the preset decarbonization standard threshold, a risk warning signal is generated; if the real-time decarbonization prediction value is greater than the preset decarbonization standard threshold, an over-limit warning signal is generated.
[0025] Specifically, the predicted value generated in step S102 is compared in real time with the preset standard threshold in step S101, and the hierarchical judgment logic is executed: Risk warning signal (Level 1): When the real-time predicted value reaches the first preset percentage (e.g., 85%) of the specified threshold. For example, if the threshold is 0.07 mm, and the predicted value reaches 0.06 mm, the system determines that there is a risk of exceeding the standard. At this time, the operating terminal will automatically pop up a yellow prompt box and push "process fine-tuning suggestions", such as suggesting fine-tuning the air-fuel ratio or shortening the current dwell time.
[0026] Exceeding Standard Warning Signal (Level 2): When the real-time predicted value exceeds the specified threshold (e.g., greater than 0.07 mm), the system determines that the steel billet has an excessive risk of decarburization. At this time, the audible and visual alarm in the control room will be triggered immediately, a red warning window will pop up on the terminal, and a "standardized handling plan" will be forcibly pushed.
[0027] Step S104: Based on the risk warning signal or the over-limit warning signal and the actual operating parameters, generate the corresponding process adjustment command and output it to the corresponding control system to perform closed-loop control of the decarburization process of the target steel.
[0028] Specifically, based on the type of warning signal and actual on-site parameters, the system generates specific process adjustment instructions to achieve closed-loop control: (1) Process intervention stage: When an over-limit warning is generated and the cumulative heating time of the billet is detected to be close to or exceed the upper limit of the threshold, the system automatically issues a "cooling / heat preservation command" to the heating furnace control system. It requires the temperature of the soaking zone to be lowered to a safe range (such as 1050-1080℃) to slow down the decarburization reaction rate and buy time for subsequent tapping.
[0029] (2) Size Rescue Stage: If, after implementing the above cooling measures, the billet still cannot be unloaded from the furnace in time due to equipment failure or production scheduling reasons, the prediction model will determine that decarburization has irreversibly exceeded the standard. At this time, the system will calculate and generate a "rolling size increment" based on the difference. For example, if the predicted decarburization exceeds the standard by 0.1 mm, the system will generate an instruction output to the rolling line control system to increase the finished product rolling diameter by 0.15 mm. By increasing the initial size of the finished product, sufficient decarburization layer removal allowance is reserved for subsequent processes (such as shot blasting and grinding), thereby avoiding the final scrapping of the product.
[0030] After each batch of production is completed, the process parameters and decarbonization detection results for that batch are collected and added to the dataset. The model parameters are slightly updated and fully retrained at preset time intervals. Simultaneously, the system records the effectiveness of each warning and response (e.g., whether exceeding limits was successfully avoided, actual decarbonization layer depth, etc.) to evaluate model performance and optimize response rules.
[0031] In addition, this application also provides a decarburization early warning and control system for the steel rolling process to implement the above method. The control system includes: a parameter and data acquisition module, a decarburization prediction module, an early warning judgment module, and a graded disposal control module.
[0032] Specifically, the parameter and data acquisition module is used to acquire the heating process parameters of the target steel and the preset decarburization standard threshold, and to collect the actual operating parameters of the target steel in the heating furnace in real time. Among them, the actual operating parameters include at least the heating temperature, cumulative heating time, and air-fuel ratio.
[0033] The decarburization prediction module is connected to the parameter and data acquisition module and is used to input the actual operating parameters into the pre-trained decarburization prediction model to generate the real-time decarburization prediction value of the target steel.
[0034] The early warning judgment module, connected to the decarbonization prediction module, is used to perform graded early warning judgments based on the real-time decarbonization prediction value and the preset decarbonization standard threshold, so as to generate a decarbonization early warning signal. The graded early warning judgment includes: if the real-time decarbonization prediction value reaches a first preset proportion of the preset decarbonization standard threshold, a risk warning signal is generated; if the real-time decarbonization prediction value is greater than the preset decarbonization standard threshold, an exceedance warning signal is generated.
[0035] The graded handling control module is connected to the early warning judgment module and the parameter and data acquisition module, respectively. It is used to generate corresponding process adjustment instructions based on risk warning signals or exceedance warning signals and actual operating parameters, and output them to the corresponding control system.
[0036] Therefore, the technical solution of this application incorporates a decarburization early warning model, which, combined with process parameter fusion and big data analysis, enables accurate early prediction of decarburization of rolled materials, changing the traditional manual experience-based judgment mode and significantly reducing the error in decarburization judgment.
[0037] In addition to the above embodiments, the present invention may have other implementation methods; all technical solutions formed by equivalent substitution or equivalent transformation fall within the protection scope claimed by the present invention.
Claims
1. A method for early warning and control of decarburization in steel rolling process, characterized in that: include: Obtain the preset heating process parameter thresholds and preset decarburization specification thresholds for the target steel grade; The actual operating parameters of the target steel in the heating furnace are collected in real time and input into a pre-trained decarburization early warning model to generate a real-time decarburization prediction value of the target steel. A graded early warning judgment is performed based on the real-time decarbonization prediction value and the decarbonization standard threshold to generate a decarbonization early warning signal. The graded early warning judgment includes: if the real-time decarbonization prediction value reaches a first preset proportion of the preset decarbonization standard threshold, a risk warning signal is generated; if the real-time decarbonization prediction value is greater than the preset decarbonization standard threshold, an over-limit warning signal is generated. Based on the risk warning signal or the over-limit warning signal and the actual operating parameters, a corresponding process adjustment instruction is generated and output to the corresponding control system to perform closed-loop control of the decarburization process of the target steel.
2. The method for early warning and control of decarburization in the steel rolling process according to claim 1, characterized in that: The training process of the decarbonization prediction model includes: Obtain historical rolling production data and corresponding offline decarburization metallographic detection data; Based on the historical rolling production data, historical heating temperature, historical heating duration, historical air-fuel ratio, and historical residual oxygen content are extracted as training feature inputs. The decarburization layer depth in the offline decarburization metallographic detection data is used as the training label output; The decarbonization prediction model is generated by training the input features and the output labels using a multivariate nonlinear regression algorithm.
3. The method for early warning and control of decarburization in the steel rolling process according to claim 1, characterized in that: The actual operating parameters include at least heating temperature, cumulative heating time, and air-fuel ratio.
4. The method for early warning and control of decarburization in the steel rolling process according to claim 1, characterized in that: The step of generating corresponding process adjustment instructions and outputting them to the corresponding control system based on the risk warning signal or the exceedance warning signal and the actual operating parameters includes: When the above-mentioned over-limit warning signal is generated, and it is detected that the cumulative heating time in the actual operating parameters reaches the upper limit of the preset time threshold and the target steel has not been taken out of the furnace, a cooling and heat preservation process adjustment command is generated. The cooling and heat preservation process adjustment command is output to the heating furnace control system to prompt the operator or automatically execute the cooling and heat preservation operations.
5. The method for early warning and control of decarburization in the steel rolling process according to claim 4, characterized in that: After outputting the cooling and heat preservation process adjustment command to the heating furnace control system, the method further includes: The updated actual operating parameters after the cooling and heat preservation operations are re-acquired and input into the decarburization prediction model to generate an updated decarburization prediction value. If the updated decarburization prediction value is still greater than the preset decarburization standard threshold, the rolling size increment is calculated based on the difference between the updated decarburization prediction value and the preset decarburization standard threshold. A rolling size adjustment command is generated based on the rolling size increment and output to the rolling line control system to adjust the finished product rolling diameter.
6. The method for early warning and control of decarburization in the steel rolling process according to claim 1, characterized in that: After generating the corresponding process adjustment command and outputting it to the corresponding control system, the process further includes: Obtain actual decarburization test data of the target steel after rolling; The actual operating parameters and the actual decarbonization detection data are added to the training dataset, and the parameters of the decarbonization prediction model are periodically iterated and optimized based on the updated training dataset.
7. A decarburization early warning and control system for a steel rolling process, used to implement the method as described in any one of claims 1-6, characterized in that, include: The parameter and data acquisition module is used to acquire the heating process parameters of the target steel and the preset decarburization standard threshold, and to collect the actual operating parameters of the target steel in the heating furnace in real time. The actual operating parameters include at least the heating temperature, cumulative heating time and air-fuel ratio. The decarburization prediction module, connected to the parameter and data acquisition module, is used to input the actual operating parameters into the pre-trained decarburization prediction model to generate the real-time decarburization prediction value of the target steel. The early warning judgment module, connected to the decarbonization prediction module, is used to perform graded early warning judgment on the real-time decarbonization prediction value and the preset decarbonization standard threshold to generate a decarbonization early warning signal. The tiered early warning judgment includes: if the real-time decarbonization prediction value reaches a first preset proportion of the preset decarbonization standard threshold, a risk warning signal is generated; if the real-time decarbonization prediction value is greater than the preset decarbonization standard threshold, an over-limit warning signal is generated. The graded handling control module is connected to the early warning judgment module and the parameter and data acquisition module, respectively. It is used to generate corresponding process adjustment instructions and output them to the corresponding control system based on the risk warning signal or the exceedance warning signal and the actual operating parameters.